New challenges for text mining: mapping between text and manually curated pathways

BMC Bioinformatics

Table 4 The dichotomy of the sentence expression for pathway representation: direct or indirect.

reaction ID	reaction type	expected annotation class	direct (annotated)	indirect (not annotated)
R1	Binding	Binding	11	56
R2	Binding	Binding	16	20
R3	phosphorylation	Protein_amino_acid_phosphorylation	19	1
+ site information	Protein_amino_acid_phosphorylation	2	5
+ modifier information	+ (any) Regulation	4	1
R4	ubiquitination	Protein_ubiquitination	5	0
R5	degradation	Protein_catabolism	27	1
	+ modifier information	+ (any) Regulation	6	3
R6	translocation	Localization	26	0
R7	Binding	Binding	23	12
R8	gene expression	Gene_expression (or Transcription + Translation)	5	9
R9	gene expression	Gene expression (or Transcription + Translation)	1	10
R10	processing	Protein_processing	5	12
		total	150	130

The sentences for pathway representation of each reaction were distributed by a direct or indirect expression. The direct expressions were annotated by ‘expected annotation class’ using GENIA event annotation, while the indirect expressions were not. The latter needs some ‘inference’ by the biological domain knowledge to construct the corresponding pathway.

ISSN: 1471-2105